Efficient Adaptive Intra Refresh Error Resilience for 3D Video Communication

Abstract

3D video wireless communication is being repeatedly called upon to play a greater role in the quest for economic and political stability, peace and understanding in the world. However, excessive error propagation in compressed 3D video can dramatically decrease perceived quality of experience (QoE). Recent error control strategies suggested that using adaptive intra refresh (AIR) tool is a good candidate for fulfilling quality of service (QoS) requirements for 3D video transmission over wireless networks. We propose multi-view video AIR (MVV-AIR) based on H.264/advance video coding (AVC) to mitigate transmission error propagation in a two-way communication system. With the MVV-AIR, both high active macroblocks and channel bit error rate (BER)are computed to generate MVV-AIR refresh map. The generated map is used to periodically insert a cyclic intra-refresh line macroblock to suppress spatiotemporal error propagation. Experimental results show that MVV-AIR outperforms traditional refresh scheme in error prone transmission environment with improved objective and subjective visual quality.